A new generation of sustainable investment (SI) indices is required to meet the needs of portfolios in today’s landscape.

It is no longer sufficient to merely improve upon a set of SI characteristics compared with a benchmark. Instead, the focus should be to achieve precise outcomes. At the same time, the desire to simultaneously control other index outcomes – such as tracking error, country/industry weights and style exposures – provides additional complications.

The ability to achieve such specific requirements in an index ultimately rests with portfolio construction – methodologies for this typically include a mechanism that permits the levels of exposure to a particular quantity to be ratcheted up and down.

Examples range from relatively simple selection and weighting techniques through to complex, optimised portfolios:

  • The former involves selecting a set of stocks with desirable characteristics and then overlaying some portfolio-weighting scheme. The selection cut-off may be used to control the levels of exposure, but not other important exposures.
  • Optimisation often does not explicitly target a precise set of exposures, but rather a composite score, tracking error or risk budget.

In contrast, the FTSE Russell tilt approach provides a transparent mechanism to exercise complete and precise control over single and multiple objectives.

A general exposure framework

The tilt methodology stems from the simple observation that one can obtain a greater degree of exposure to, say value, than is possible via a given benchmark by multiplying that benchmark’s weights by a positive score, which varies monotonically with the value characteristic of individual stocks.

In this equation, an initial set of benchmark weights multiplied by a set of positive numbers representing the value scores equals the set of “tilted” portfolio weights (one for each stock). This results in a set of unadjusted stock weights that require normalisation (or rescaling) to be equal to one.

FTSE Russell uses this convention in equations throughout its research paper "Targeted Sustainability", where it is understood that the normalisation is implicit.

Each exposure target merely requires incorporating an additional term in the multiplication formula. All exposure objectives are, therefore, embedded in exactly the same way. Indeed, additional implementation properties – such as levels of investment capacity, maximum stock weights and turnover – may also be expressed and controlled as additional multiplication tilts.

Recall that the exponents (or tilt strengths) in this formula determine the degree of the exposure to each metric. Therefore, given a target for each exposure, it is possible to determine the set of tilt strengths required to achieve each exposure target.

This, in essence, is how FTSE Russell constructed the “Target Exposure Index”.

The flexibility of this approach enables a variety of sustainable indices to be created with different use cases. Some may concentrate on reducing carbon emissions and reserves, while neutralising style exposures and controlling tracking error; others may focus on smart sustainability, combining climate considerations and active style exposures.

ESG and low carbon indices

The FTSE Russell paper focuses on portfolios with improved ESG ratings along with lower carbon emissions and reserves than the market capitalisation benchmark. Other SI objectives covering the exclusion of companies involved in contentious product activities – such as weapons, tobacco, thermal coal and nuclear power – or controversies related to the UN Global Compact principles are additionally incorporated.

Further, the aim is to achieve this without excessive tracking error or by skewing the resulting index to particular countries or industries.

More precisely, assume we require an increase in ESG ratings of 20% and reductions in both emission and reserve intensities of 50%, relative to the benchmark. We control tracking error as well as country and industry weights by imposing country neutrality and an active industry weight constraint of +/-5%. This can be achieved by constructing a multiple tilt index with tilts to improved ESG, low emissions and low reserves, with additional tilts to control country and industry weightings.

The Target Exposure methodology is an evolution of FTSE Russell’s tilting methodology, whereby tilt strengths are chosen to obtain a given set of exposure targets. This fits perfectly with the increasing demand for index solutions, which incorporate precise low carbon and ESG outcomes.

In its paper, FTSE Russell provides a thorough examination of a simulated index, which is the result of the construction process generally described earlier in this article.

The simulated index includes additional tilts that limit the maximum stock weight to the minimum of 10% and 10 times the market capitalisation weight. An exclusion list is also to remove companies involved with weapons, tobacco, thermal coal and nuclear power, in addition to companies involved in controversies related to the UN Global Compact Principles. The controversies exclusions list is re-applied at each subsequent quarter. Stock weights of less than 0.5 basis points are also removed.

FTSE Russell also examines simulated results for several FTSE universes – the outcomes of which are shared in the paper’s appendix. They have simulated indices that employ the target exposure framework to combine precise SI objectives targeting the control of important portfolio characteristics such as country and industrial weightings, levels of diversification and capacity.

Such indices, therefore, not only provide precisely targeted SI solutions, but are also designed for ease of implementation and limited levels of tracking error.


Read the full research paper here


Authors from FTSE Russell: Andrew Dougan, director, research and analytics; Fong Yee Chan, senior product manager, sustainable investment; and Elizeveta Nebykova, senior analyst.